RLE Encoding...What's the wrong?

Question

I'm trying to make a RLE (Run-Length Encoder) Programme just for characters. I read the way it works on notes on net. And I tried to fix my code! Regardless I think that the steps of code are right, the code doesnt work! It appears some strange 'Z' as it runs. I really can't find what;s wrong! Could you please give me a piece of advice?

#include <stdio.h>

int main()
{
    int count;
    unsigned char currChar,prevChar=EOF;
    while(currChar=getchar() != EOF)
    {
        if ( ( (currChar='A')&&(currChar='Z') ) || ( (currChar='a')&&(currChar='z') ) )
        {
            printf("%c",currChar);
            if(prevChar==currChar)
            {
                count=0;
                currChar=getchar();
                while(currChar!=EOF)
                {
                    if (currChar==prevChar)
                        count++;
                    else
                    {   
                        if(count<=9)
                            printf("%d%c",count,prevChar);
                        else
                        {   
                            printf("%d%c",reverse(count),prevChar);
                        }                       
                        prevChar=currChar;
                        break;
                    }
                }
            }
            else
                prevChar=currChar;

            if(currChar==EOF)
            {   printf("%d",count);
                break;
            }
        }
        else
        {
            printf("Error Message:Only characters are accepted! Please try again! False input!");
            break;
        }
    }
    return 0;
}

int reverse(int x)
{
    int p,y,r=0;
    x=(x<0)?-x:x;
    while (x>0)
    {
        y=x%10;
        p=x/10;
        r=10*r+y;
        x=p;
    }
    printf("%d",r);
    return 1;
}

e.g. I gave the input:

AAAAAAAAAAAAAAAAAAAAFFFFFFFFFFFFFFFFFFFFFFFFFFEEEEEEEEEEEEEEEEEEEEEEEEEGGGGGGGGGGGGGGGGVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVVSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSRRRRRRRRRRHD RRRRRRRRRRRRRRRRRRRRHHHHHHHHHHHHHHHHAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYNNNNNNNNNNNNNNNNNNNNNNNNNNNNMMMMMMMMMMMMMMMMMMMMMMMMMMMMS TTTTTTTTTTHHHHHHHHHHHH

and I got the output:

Z0AZZ0AZZ0FZZ0FZZ0FZZ0FZZ0FZZ0FZZ0FZZ0FZZ0FZZ0EZZ0EZZ0EZZ0EZZ0EZZ0EZZ0EZZ0EZZ0GZZ0GZZ0GZZ0GZZ0GZZ0GZZ0VZZ0VZZ0VZZ0VZZ0VZZ0VZZ0VZZ0VZZ0VZZ0VZZ0SZZ0SZZ0SZZ0SZZ0SZZ0SZZ0SZZ0SZZ0SZZ0SZZ0SZZ0SZZ0RZZ0RZZ0RZZ0RZZ0 ZZ0RZZ0RZZ0RZZ0RZZ0RZZ0RZZ0HZZ0HZZ0HZZ0HZZ0HZZ0HZZ0AZZ0AZZ0AZZ0AZZ0AZZ0AZZ0AZZ0AZZ0AZZ0AZZ0

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! :(

Answer 1

Look at this line:

if ( ( (currChar='A')&&(currChar='Z') ) || ( (currChar='a')&&(currChar='z') ) )

you are assigning 'A' to currChar then you are assigning 'Z' to currChar and so on...

You need to change = to == to make it comparison instead of assignment.

Also, what do you mean by (currChar='A')&&(currChar='Z')? currChar cannot be 'A' and 'Z' at the same time, I believe what you meant to put here was a check for currChar being included in certain interval. So it should probably be:

(currChar>='A')&&(currChar<='Z')

The same applies for the second part of your condition.

Answer 2

Converting previous comments into an answer...

• There are problems with assignments in the 'if' conditions - as pointed out in the other answer.

• You cannot assign a value to an unsigned char and then expect to detect EOF. Forget the name - remember that getc() and getchar() (and fgetc()) return an integer, not a char; they have to return an integer because they need to return every possible valid character value plus EOF!

• Your test at if(currChar=='EOF') is bizarre. You are using a multi-character constant, which is at best implementation-defined, and which is not going to equal EOF (unquoted) as returned by getchar(). Plus the type of currChar is wrong.

• reverse() always returns 0; is that what you really wanted?

• The line 'while(currChar=getchar() != EOF)' needs extra parentheses so that it works as expected: 'while((currChar = getchar()) != EOF)'. At the moment, it assigns 0 (NUL, '\0') or 1 (Control-A) to currChar.

• The inner while loop does not read any characters, so it is going to send your program into a frenzy - it needs the assigning 'while((currChar = getchar()) != EOF)' notation too. Then you need to work out what the expected output really is, because I don't think it makes much sense - in particular, the stray 'printf("%c",currChar);' after the main test is dubious - maybe it is a debugging print that you left behind by accident.

• You also need to consider how the code should handle things like newlines - and that's before we get to issues of ambiguity in the output (how do you tell the difference between the RLE encoded data and data containing numeric values. There is a lot to worry about in your algorithm! It is mostly not correct, I'm sorry to report.

Here is some semi-working code; it explicitly refuses to deal with digits (but that's all).

/* RLE - Run Length Encoding */
/* SO 2485285 */

/*
** Input:  stream of data except for digits 0-9
** Output: stream of data with adjacent sets of 3 or more of the same
**         character represented by 3Z (for ZZZ), etc.
*/

#include <stdio.h>
#include <ctype.h>

static void print_rle(int count, int repchar)
{
    if (count > 2)
        printf("%d%c", count, repchar);
    else if (count == 2)
        printf("%c%c", repchar, repchar);
    else if (repchar != EOF)
        printf("%c", repchar);
}

int main()
{
    int count = 1;
    int currChar;
    int prevChar = EOF;

    while ((currChar = getchar()) != EOF)
    {
        if (isdigit(currChar))
            fprintf(stderr, "Bogus character %c read - ignored\n", currChar);
        else if (currChar == prevChar)
            count++;
        else
        {
            print_rle(count, prevChar);
            count = 1;
            prevChar = currChar;
        }
    }
    print_rle(count, prevChar);

    return 0;
}

And this is the output when I run it on its own source code (note that I use spaces, not tabs). The 'Bogus character' messages are printed on stderr, not stdout.

/* RLE - Run Length Encoding */
Bogus character 2 read - ignored
Bogus character 4 read - ignored
Bogus character 8 read - ignored
Bogus character 5 read - ignored
Bogus character 2 read - ignored
Bogus character 8 read - ignored
Bogus character 5 read - ignored
/* SO  */

/*
Bogus character 0 read - ignored
Bogus character 9 read - ignored
** Input:  stream of data except for digits -
Bogus character 3 read - ignored
** Output: stream of data with adjacent sets of  or more of the same
Bogus character 3 read - ignored
**9 character represented by Z (for 3Z), etc.
*/

#include <stdio.h>
#include <ctype.h>

static void print_rle(int count, int repchar)
{
Bogus character 2 read - ignored
4 if (count > )
8 printf("%d%c", count, repchar);
Bogus character 2 read - ignored
4 else if (count == )
8 printf("%c%c", repchar, repchar);
4 else if (repchar != EOF)
8 printf("%c", repchar);
}

int main()
{
Bogus character 1 read - ignored
4 int count = ;
4 int currChar;
4 int prevChar = EOF;

4 while ((currChar = getchar()) != EOF)
4 {
8 if (isdigit(currChar))
12 fprintf(stderr, "Bogus character %c read - ignored\n", currChar);
8 else if (currChar == prevChar)
12 count++;
8 else
8 {
12 print_rle(count, prevChar);
Bogus character 1 read - ignored
12 count = ;
12 prevChar = currChar;
8 }
4 }
4 print_rle(count, prevChar);

Bogus character 0 read - ignored
4 return ;
}